NVRAM: read failed

[jhigham]

Running ESXi 4.0. The past two nights the box has locked up and locked up all VM's. I had very limited access to the host via the windows client, and no access to the VM's. I was able to see the last entry in the host event log ‘NVRAM: read failed'. This entry was tied/linked to one of the VM's. I do not know if this was the same error as previous times. Any and all help/suggestions are appreciated.

[jhigham]

Problem resurfaced last night. Same error, same virtual machine. Host was unresponsive and required reset button to be pressed. Is it something specific to that VM or is it a bigger issue with the host?

[DSTAVERT]

You should post a little more information to help forum members understand the problem.

You could be experiencing disk problems, RAM problems. It could be just an issue with a corrupt NVRAM file. Post some more information. What led up to the problem. Hardware, Software versions, etc etc.

[J1mbo]

Can you list the hardware spec of the host?

I'd start by running a thorough memory test on the box, for 24 hours.

Please award points to any useful answer.

[jhigham]

ESXi 4.0, upgraded from 3.5

1X - Intel S5000PSL

2X - Intel 2.5 GHz Quad Core Xeon

4X - Crucial 4GB, 240-pin DIMM, DDR2 PC2-5300 memory module

1X - Adaptec RAID 5805, running 3 arrays, each in RAID 6

8X - Seagate 750GB SATA

2X - Intel Dual Port NIC

1X - Intel RMM

Adaptec RAID controller is not reporting any issues with the arrays. They are showing as optimal. I have not yet run memtest.

The VM that is displaying the NVRAM error is CentOS running cPanel as a DNS server.

[KenzoKono]

Hi, I have been facing same problem on my ESXi4. It was up and running almost 4 month without any reboot with iSCSI on Windows Storage Server 2008. Problem happend the day before yesterday and I sawed the same messages only on Windows Servers. The other linux system looks OK, but the ESXi Hyper Visor was getting too slow likes over heated. So I decided to do a hardware reset and after 1 day, same problem happened again.

This is the action I did. I did windows update on Windows Storage Server 2008 and rebooted it once all patches were applied

It looks working fine now, and will give a post when I face same problem again.

[DSTAVERT]

I can't imaging the nvram file causing the host to become unstable but I suppose anything is possible. The nvram file can be deleted with no ill effects. It is the VM BIOS settings. It will be recreated when the VM is restarted.

I would point the vmware logs to another location so that they are retained after reboot. You can have ESXi write logs to a directory on a datastore. You can do that from the Client in Configuration / Software / Advanced / Syslog.